A publisher app appears in your placement report, someone calls it wasted spend, and a blocklist arrives before anyone checks what the report measures. For an app acquisition team, that sequence can turn a legitimate suitability concern into an unsupported performance claim—and an account-wide change into a surprise for other campaigns.
Google App campaigns can respect account-level exclusions for unwanted websites and apps. Before adding them, identify the exact placement, confirm the applicable account or manager scope, and record why it should be excluded. App campaign placement reports currently provide impressions, not placement-level cost or conversions, so they cannot establish a publisher's CPA or the savings from blocking it. Google's network-distribution guidance identifies account-level placement exclusions as the route for unwanted sites or apps; its placement-report documentation explains the reporting limits.
This guide is for founders, UA leads and growth teams deciding whether to approve an exclusion, ask for better evidence, or commission a broader account review. The decision register and acceptance workflow below are original editorial guidance, rather than a Google policy or a claim about private client results.
First distinguish the app you advertise from the app carrying the ad
“Exclude mobile apps” can describe two very different jobs. A retailer running Display may want to avoid showing ads inside publisher apps. A mobile-app company running App campaigns may want to stop its acquisition ads appearing in a particular publisher's inventory. A tutorial for the first job does not establish the controls available for the second.
Write a request that names both sides: “For the campaign promoting our budgeting app, investigate the reported placement associated with this publisher app.” That is sufficiently precise to investigate. “Remove app traffic” leaves the campaign type, inventory and intended business effect unresolved.
Google's assets and ads documentation describes automated placement selection and states that App campaigns respect account settings for webpage and app exclusions. This supports investigating a specific exclusion. It does not turn App campaigns into manually selected publisher buying.
Also separate the object being excluded. A publisher, search query, country and audience are different objects. If the concern is search intent, use the Google App campaigns negative-keyword framework. If the concern is suspicious acquisition activity, preserve evidence for an app install-fraud investigation. A publisher's name in an impression report does not prove fraud.
What the placement report can actually establish
Use the report to identify observed exposure and prepare a suitability review. Do not begin with a spreadsheet column called “wasted budget” unless an independently valid dataset supports that calculation.
Google documents a predefined App campaigns placement report in Report editor and a custom report built using the placement attribute and impressions. It also explains that lower-activity placements can be consolidated for privacy, so granular totals can fall below campaign totals. Standard campaign reports remain the reference for billing and total delivery. Some Play Store video placements may appear under Search in network segmentation. Google Ads placement reporting describes these limitations.
The practical consequence is a hierarchy of evidence. A visible row supports the statement that the report recorded impressions for that placement during the selected period. It does not identify the users acquired there, their downstream contribution, or the counterfactual outcome if that placement had been unavailable.
Keep the original export, date range, time zone, campaign filters and extraction time. Record whether the request concerns one observed row or a wider class of content. Without that distinction, an investigation can quietly expand from reviewing one placement to blocking a large category of potentially useful inventory.
Treat missing rows carefully, too. A placement disappearing from a later report is consistent with reduced visible exposure, but visibility limits and changing delivery make that an incomplete causal test. State exactly what was observed and leave the stronger conclusion open.
Match the concern to the right decision
Concern Evidence to request Appropriate next decision --- --- --- Specific publisher conflicts with an approved suitability rule Placement identity and documented rule Review a narrowly scoped exclusion Publisher is called expensive A valid cost and outcome dataset with compatible definitions Investigate measurement before claiming a saving Team wants to remove an entire network Campaign subtype, available controls and commercial requirement Assess product fit; do not substitute a publisher list for network selection Large inherited blocklist has no owner Current associations, reasons and change history Audit governance before expanding the list Suspected invalid acquisition activity Independent integrity evidence and attribution definitions Investigate fraud separately from suitability
A business can reasonably decide that a placement is unacceptable without proving that it performs badly. The record should say that the exclusion enforces an approved suitability standard. That is a defensible governance decision. Inventing an efficiency story makes the same decision harder to evaluate later.
Conversely, an unfamiliar publisher name is not enough to justify a permanent performance exclusion. Ask what evidence would change the recommendation. If there is no answer, the proposal is an untested preference. It may still deserve review, but it should not be sold internally as an optimization result.
Use a four-field exclusion register
Every proposed entry should have an identity, a reason, a scope and an acceptance condition. This small register makes the decision reviewable without turning it into a lengthy compliance exercise.
Identity: retain the exact reported domain, app identifier, channel or video and the evidence linking it to the concern. Human-readable names can be ambiguous. Do not paste a publisher's marketing website into an app exclusion field and assume the two objects are interchangeable.
Reason: choose a specific category such as documented brand suitability, contractual restriction, or a performance hypothesis requiring more evidence. Attach the relevant business rule or analysis. “Low quality” is too vague to distinguish a policy requirement from an analyst's opinion.
Scope: name the customer account, manager relationship, relevant campaigns and placement type. Identify who else must review the change. The desired campaign may be only one consumer of a broader account setting.
Acceptance: define what can be verified after the change. A saved setting, an account-level change-history entry and a later exposure review are different checks. A lower blended CPA is a separate business observation and cannot substitute for verifying that the intended setting was applied.
Add an owner and review date to the register. A mandatory business restriction may need ongoing enforcement; a speculative performance exclusion should have a deliberate reconsideration point. Preserve the reason for retaining or removing an entry rather than letting old blocklists become unexplained infrastructure.
Verify scope before touching a shared list
Account and manager settings deserve particular care when an agency operates multiple brands or business units. Google's manager-account exclusion-list guidance describes applying lists to selected client accounts and warns that list changes are reflected in every associated account. Its instructions also distinguish these account-level associations from campaign-specific application.
Before editing, obtain the actual list associations. A screenshot of a list name is not a scope map. The reviewer should be able to answer which customer accounts inherit the list, who maintains it, and whether another team relies on an entry or exception.
For example, a publisher unsuitable for one regulated product may be acceptable for another app. That does not imply either team is wrong. It means a shared restriction needs a common approved rule, or the operating structure needs review before the change is approved.
Do not create a new account or restructure campaign ownership merely to make a proposed exclusion easier. Start with the controls and boundaries actually present. If a requirement cannot be expressed safely at the available level, document the limitation and discuss a supported alternative with the account owner or Google support.
The signoff should explicitly acknowledge other affected campaigns. A change approved as a single-app optimization can otherwise become an unreviewed restriction on another product's acquisition. This is why the scope field belongs in the brief before implementation, rather than being added to the report afterward.
Treat formatting as part of the control
A valid business reason can still produce the wrong setting if the identifier is malformed. Google's account-level placement exclusion guide distinguishes bare domains from entries retaining www, requires separate country domains, and documents different subdomain behavior on Search partners. It also notes that Google-domain URLs cannot simply be added as direct placement exclusions.
Use the current official instructions for the exact object and network. Do not normalize every entry with an indiscriminate spreadsheet formula. Removing a prefix may intentionally broaden a domain exclusion; retaining it may narrow the affected property. Both choices need to match the approved scope.
A file can be syntactically accepted while still representing the wrong business object. Have a second reviewer compare a small set of intended entries with their source evidence before a larger import. Confirm what the list will affect, rather than checking only whether the upload succeeds.
Avoid borrowed “universal” blocklists whose rationale cannot be inspected. The relevant question is whether an entry belongs in this account under its approved rules. The popularity or length of a third-party list does not establish suitability, fraud, or economic value for your app.
Implement with a narrow verification sequence
Prepare the baseline first: current settings, list associations, a placement export and standard campaign reports covering the relevant period. Preserve enough context to understand subsequent delivery changes without retaining unnecessary personal data or unrelated account exports.
Then follow a bounded sequence:
Obtain approval for the reason and affected account scope.
Validate the exact identifier and the relevant official instructions.
Apply the smallest supported change that expresses the requirement.
Confirm the saved setting and record the change-history evidence.
Review later reporting using the same filters and documented limitations.
Assess campaign-level business outcomes separately, recording concurrent changes.
Google's account guide describes where exclusions can be entered and how account-level Change history can confirm the edit. Its timing guidance is described as typical, not a delivery guarantee. Use the current official instructions, and avoid promising that every report will immediately become a complete verification record.
If the setting is unavailable, rejected or behaves unexpectedly, preserve the observed state and escalate that specific question. A useful support request includes the campaign subtype, intended placement identifier, setting level and redacted before-and-after evidence. Repeatedly importing larger lists is not a substitute for understanding the failure.
A fictional example: exposure is not wasted spend
Consider an invented app campaign with 100,000 campaign impressions over a review period. Its placement export identifies 60,000 impressions in named rows, including 12,000 for a publisher under review. These numbers are illustrative arithmetic, not a benchmark or client result.
That publisher represents 12% of campaign impressions and 20% of the named-row subtotal. Both percentages can be correct because they use different denominators. Neither means the publisher consumed 12% or 20% of budget.
Suppose total campaign cost was 8,000 units of account currency. Multiplying that cost by either exposure percentage would create an unsupported placement-cost estimate. Without compatible placement-level cost evidence, the resulting “saving” has no defensible basis. Attributed conversions and downstream revenue cannot be allocated by impression share either.
Now suppose the publisher clearly violates a documented suitability requirement. The team can approve the exclusion for that reason, preserve the scope evidence and verify implementation. It need not invent a financial estimate to justify the decision.
If campaign CPA later changes, record it alongside budget, bid, creative, audience and market changes. The observation can motivate further analysis, but a before-and-after movement does not isolate the exclusion's effect. The business may have made a sound suitability decision while remaining unable to quantify its incremental financial impact.
What a useful agency audit should hand back
An audit should produce a decision record the account owner can act on. Ask for the proposed exclusion register, verified associations, source exports, remaining unknowns and a clear separation between mandatory restrictions and hypotheses.
The handoff should name who can approve and apply changes, how another affected team will be notified, and what evidence closes each verification step. It should also explain when the recommendation is to leave settings alone pending better evidence. A longer blocklist is not inherently a better audit deliverable.
If creative coverage or campaign goals are the underlying concern, exclusions may be an indirect and poorly evidenced intervention. Review the App campaign creative assets guide before using inventory restrictions to compensate for missing creative. Google describes App campaign distribution as automated within campaign inputs; a list should not be presented as a guaranteed way to force a desired channel mix. Google Ads network-distribution guidance provides the relevant context.
Questions to settle before approval
Can we calculate placement CPA from the App placement report? No. Its documented performance metric is impressions. Any placement-level economic claim needs another valid source with matching scope and definitions.
Does excluding an app mean excluding every app placement? No. A specific publisher identifier and a network-wide restriction are different requests. Verify the supported setting rather than interpreting the label loosely.
Will exclusions lower acquisition costs? That outcome cannot be promised. A restriction removes opportunities; the resulting economics depend on delivery, measurement and user value. Suitability requirements can justify action independently of a savings claim.
What if the evidence is incomplete? Approve only what the available evidence supports. Keep unresolved financial hypotheses separate, document the limitation, and choose a review or support action with a named owner.
Bring the evidence to an app acquisition conversation
Sharply Labs' mobile app growth work and paid acquisition services are relevant to established teams that need to separate publisher suitability, inherited account restrictions and unsupported performance conclusions.
For an initial conversation, bring a redacted placement export, the current exclusions and a map of relevant manager relationships. The conversation can establish which decision needs evidence, who owns the affected settings and what validation work should be scoped. A detailed account audit would be agreed separately. No complete placement visibility, exclusion-driven savings, acquisition-cost reduction or campaign result is promised.